What is the L in numpy.shape and 32 in numpy.type of an array?

Question

I'm trying the functions of numpy arrays and below is the code:

import numpy as np

Z =np.array( 
    [[0,4,0,0,0,0],
     [0,0,0,1,0,0],
     [0,1,0,1,0,0],
     [0,0,1,1,0,0],
     [0,0,0,0,0,0],
     [0,0,0,0,0,0]])
print Z
print Z.dtype
print Z.shape

Which gave:

[[0 4 0 0 0 0]
 [0 0 0 1 0 0]
 [0 1 0 1 0 0]
 [0 0 1 1 0 0]
 [0 0 0 0 0 0]
 [0 0 0 0 0 0]]
int32
(6L, 6L)

It is an integer array with 6 rows and 6 columns. But what is the 32 from numpy.type and the L in numpy.shape?

Answer 1

The 32 in the array type refers to 32-bit width of the underlying machine integer that stores the array. This means that an array with, say, 1 million elements, will take up at least 4 million bytes (32 million bits) of memory.

On the other hand, the suffix L returned by the shape property has nothing to do with the data you can put in your arrays and you should not be concerned with it. If you are interested in the techical details, L denotes the long Python type used to represent integers of unbounded width (not to be confused with the C type of the same name). It doesn't make much sense to represent a small integer such as the number 6 as long, but some code does it anyway for consistency with the same API returning a larger integer. For example, the os.stat call always returns byte sizes in long integers, even if they would fit the regular int type, in order to maintain type invariance of its return value.